Competitive 1,2- and 1,5-hydrogen shifts following 2-vinylbiphenyl photocyclization

Article abstract of DOI:10.1021/jo051730y

The photocyclization of 2-vinylbiphenyl and its derivatives has been proposed to occur via a two-step mechanism: photocyclization to form an unstable 8a,9-dihydro-phenanthrene intermediate, followed by exothermic unimolecular isomerization to a 9,10-dihyd

Full text of DOI:10.1021/jo051730y

Competitive 1,2- and 1,5-Hydrogen Shifts Following
2-Vinylbiphenyl Photocyclization
Frederick D. Lewis,*,† Meledathu C. Sajimon,† Xiaobing Zuo,‡ Michael Rubin,§,| and
Vladimir Gevorgyan§
Department of Chemistry, Northwestern University, Evanston, Illinois 60208-3113, Chemistry Division,
Argonne National Laboratory, Argonne Illinois 60439, and Department of Chemistry, University of
Illinois at Chicago, Chicago, Illinois 60607-7061
lewis@chem.northwestern.edu
Received August 15, 2005
The photocyclization of 2-vinylbiphenyl and its derivatives has been proposed to occur via a two-
step mechanism: photocyclization to form an unstable 8a,9-dihydro-phenanthrene intermediate,
followed by exothermic unimolecular isomerization to a 9,10-dihydrophenanthrene. The mechanism
of the hydrogen shift process has been investigated using deuterated derivatives of 2-isopropenyl-
biphenyl and 2,6-diphenylstyrene. 1H NMR analysis of the photoproducts indicates that the
thermally allowed 1,5-hydrogen or deuterium shift is a minor product-forming pathway and that
an unusual double 1,2-hydrogen or deuterium shift is the major product-forming pathway. The
potential energy surface for photocyclization and hydrogen shift processes has been explored
computationally. The calculated barrier for the 1,5-shift is predicted to be significantly lower than
that for the 1,2-shift. Alternative mechanisms for the occurrence of 1,2-hydrogen or deuterium
migration are presented.
SCHEME 1
Introduction
The thermal sigmatropic 1,5-hydrogen shift reactions
of cyclic dienes provide textbook examples of pericyclic
reactions.1 In contrast to the extensively documented
reactions of cyclopentadienes, the analogous reactions of
1,3-cyclohexadienes have received relatively little atten-
tion. An investigation of the rearrangement of 1,4-d2-
cyclohexadiene provided a value of Ea ) 41 kcal/mol,2
substantially higher than that for cyclohexadiene, 24.3
kcal/mol.3 Density functional calculations have repro-
duced the higher experimental barrier for cyclohexadiene,
which is attributed to added strain in the transition state,
which has both a larger H‚‚‚C‚‚‚H angle and longer C1-
C5 distance when compared to the transition state for
cyclopentadiene.4,5
Our interest in the thermal rearrangement of cyclo-
hexadienes was stimulated by recent studies of the
photochemical reactions of 2-vinylbiphenyls.6,7 The pho-
tochemical conversion of 2-vinylbiphenyl 1 to 9,10-
dihydrophenanthrene 1a (Scheme 1) was initially de-
scribed by Horgan et al.8 They proposed a two-stage
† Northwestern University.
‡ Argonne National Laboratory.
(4) Hess, B. A.; Baldwin, J. E. J. Org. Chem. 2002, 67, 6025-6033.
(5) Alabugin, I. V.; Manoharan, M.; Breiner, B.; Lewis, F. D. J. Am.
Chem. Soc. 2003, 125, 9329-9342.
§ University of Illinois at Chicago.
| Current address: Department of Chemistry, University of Kansas,
Lawrence, KS 66045-7609.
(6) Lewis, F. D.; Zuo, X.; Gevorgyan, V.; Rubin, M. J. Am. Chem.
Soc. 2002, 124, 13664-13665.
(1) Woodward, R. B.; Hoffmann, R. The Conservation of Orbital
Symmetry; Verlag Chemie: Weinheim, 1970.
(2) De Dobbelaere, J. R.; Van Zeeventer, E. L.; De Haan, J. W.; Buck,
H. M. Theor. Chim. Acta 1975, 38, 241-244.
(3) Roth, W. R. Tetrahedron Lett. 1964, 1009-1013.
(7) Lewis, F. D.; Zuo, X. Photochem. Photobiol. Sci. 2003, 2, 1059-
1066.
(8) Horgan, S. W.; Morgan, D. D.; Orchin, M. J. Org. Chem. 1973,
38, 3801-3803.
10.1021/jo051730y CCC: $30.25 © 2005 American Chemical Society
Published on Web 11/09/2005
J. Org. Chem. 2005, 70, 10447-10452
10447

Lewis et al.
SCHEME 2
SCHEME 3
The photochemical behavior of 1-3 has been re-
ported.6,7 Irradiation at 298 K in cyclohexane solution
results in quantitative conversion to the cyclized isomers
1a-3a (Scheme 1). The quantum yields for formation of
3a and 3a-d10 at 298 K are identical. Irradiation of 3 at
77 K in a methylcyclohexane (MC) glass results in the
formation of a 535 nm absorption band assigned to the
intermediate 3i.6 No long-wavelength absorption band is
observed upon irradiation of 1 in a MC glass at 77 K,
and a very weak band is observed for 2, in accord with
the low populations of their reactive syn rotamers at low
temperature.7 The 535 nm band assigned to 3i disappears
rapidly upon warming of the glass, and it is not observed
when either 3 or 3-d10 are irradiated at 90 K in liquid
propane.
We have also attempted to study the formation and
decay of 3i by means of femtosecond time-resolved
pump-probe spectroscopy of 3 at room temperature. A
weak transient with an absorption maximum at 510 nm
is formed within the ca. 1 ps rise time of the exciting 267
nm laser pulse and decays with a lifetime longer than
the 100 ps probe period. This result is consistent with
an earlier report by Fornier de Violet et al. of the
observation of a 510 nm intermediate with a decay time
of ca. 1 ms upon laser flash photolysis of a 2-vinylbi-
phenyl derivative.11
mechanism for product formation: photocyclization to
yield the unstable 8a,9-dihydrophenanthrene 1i followed
by a thermal intramolecular hydrogen shift to yield 1a.
As a consequence of the gain in aromaticity, the 1,5-
hydrogen shift is highly exothermic. Alabugin et al. have
calculated values of ∆Hrxn ) -54 kcal/mol and Ea ) 17
kcal/mol for the 1,5-hydrogen shift of 1i.5
The photocyclization of 1 and other vinylbiphenyls is
subject to ground state conformational control, cyclization
occurring from the minor conformer syn-1 but not from
the major conformer anti-1 (Scheme 1).7 Since the singlet
anti conformer is unreactive, the quantum yield for the
formation of 1a is low (<0.01), even though the prepara-
tive yield is quantitative. An increase in the ground-state
population of syn-2 is accompanied by an increase in the
quantum yield for cyclization. 2-Vinyl-1,3-terphenyl 3
undergoes photocyclization to yield 3a in high quantum
yield upon irradiation in fluid solution, as a consequence
of symmetry-enforced conformational control.6 Irradiation
of 3 at 77 K in a methylcyclohexane glass results in the
formation of a dark purple product assigned to the
intermediate 3i. However, upon warming the glass to 120
K, the color disappears within a few seconds, indicative
of a thermal barrier much lower than the value calculated
for 1i.9
A long wavelength intermediate is observed upon
irradiation of the conformationally constrained 2-vinyl-
biphenyl derivative 4 in liquid propane at low tempera-
tures (Scheme 3). The first-order kinetics for thermal
bleaching of the absorption band assigned to 4i over the
temperature range 90-130 K provides values of Ea ) 2.5
kcal/mol and log A ) 3.3, which are consistent with a
tunneling mechanism.12
To obtain further information about the rearrangement
mechanism, we have investigated the photochemical
1
behavior of deuterated derivatives of 2 and 3. H NMR
spectral analysis of the photocyclization products ob-
tained from these derivatives indicates that a double 1,2-
hydrogen shift is the major pathway for product forma-
tion at room temperature. Barriers for the 1,5- and 1,2-
hydrogen shifts and ring-opening isomerization of 2i have
been calculated at the B3LYP/6-311G**//B3LYP/6-31G**
level and provide the basis for discussion of the isomer-
ization mechanism.
Deuterium Labeling Studies. The upfield region of
1
the H NMR spectra of the products obtained following
irradiation of 2 and 2-d5 in cyclohexane solution are
shown in Figure 1. The spectrum of 2a has been assigned
by Harvey et al. and is consistent with the chemical shifts
and coupling constants.13 Comparison of the spectrum of
the product obtained from 2-d5 with that from 2 indicates
that aromatization of the putative intermediate 2i occurs
predominantly via a 1,2-hydrogen migration rather than
a 1,5-hydrogen migration (Scheme 4a). A ratio of ca. 90:
10 for 1,2- vs 1,5-migration products can be calculated
from the integrated peak areas for 2a-d5.
Results and Discussion
Synthesis and Photochemical Behavior. The syn-
theses of 2 and 3 and their deuterated analogues 2-d5
and 3-d10 are outlined in Scheme 2. Reaction of 2-biphe-
nyl-magesium bromide and acetone or acetone-d6, fol-
lowed by acid-catalyzed dehydration, affords 2 and 2-d5,
respectively. The synthesis of 3 via the palladium-
catalyzed homodimerization of 1-(but-3-en-1-ynyl)ben-
zene has been reported.10 Reaction of the d5-phenyl
analogue provided 3-d10. Synthetic procedures and prod-
uct characterization are described in Experimental Sec-
tion.
The upfield region of the 1H NMR spectra of solutions
of 3 and 3-d10 obtained following irradiation in cyclohex-
ane solution are shown in Figure 2. 1H NMR assignments
for the two methylene multiplets of 3a proved essential
to the interpretation of the deuterium labeling studies.
(11) Fornier de Violet, P.; Bonneau, R.; Lapouyade, R.; Koussini,
R.; Ware, W. R. J. Am. Chem. Soc. 1978, 100, 6683-6687.
(12) Sajimon, M. C.; Lewis, F. D. Photochem. Photobiol. Sci. 2005,
4, 789-791.
(13) Harvey, R. G.; Fu, P. P.; Rabideau, P. W. J. Org. Chem. 1976,
41, 3722-3725.
(9) Zuo, X. Ph.D. Thesis, Northwestern University, 2002.
(10) Gevorgyan, V.; Tando, K.; Uchiyama, N.; Yamamoto, Y. J. Org.
Chem. 1998, 63, 7022-7025.
10448 J. Org. Chem., Vol. 70, No. 25, 2005

Mechanism of 2-Vinylbiphenyl Photocyclization
FIGURE 1. 1H NMR spectra (CDCl3) of the products 2a and
2a-d5 obtained from irradiation of 2 and 2-d5, respectively, in
cyclohexane solution at room temperature.
FIGURE 2. 1H NMR spectra (CDCl3) of the products 3a and
3a-d10 obtained from irradiation of 3 and 3-d10, respectively,
in cyclohexane solution at room temperature.
SCHEME 4
1,2-shifts have previously been proposed by Rodr´ıguez
et al. as an alternative to a 1,5-shift for the intramolecu-
lar isomerization of isobenzene (1,2,4-cyclohexatriene) to
benzene.14 However there is no experimental evidence for
the occurrence of either process.
Barriers for the 1,5-shift of 1i and 3i have been
calculated by Alabugin et al. at the B3LYP/6-31G**
level.5 The values they report are 16.6 and 16.2 kcal/mol.
The barriers obtained using B3LYP density functional
theory are somewhat larger than the experimental values
for 1,5-shifts in five-membered rings but are in better
agreement with experimental values for six-membered
rings.5 We have calculated the barriers for the 1,5-
and 1,2-shifts and the ring-opening isomerization of 2i
at the B3LYP/6-311G**//B3LYP/6-31G** level using
GAUSSIAN 98.15 The results are shown in Scheme 5
along with the singlet energies of 2 and 2i obtained from
the 0,0 bands of their absorption spectra. The calculated
barrier for the 1,5-shift is 18.3 kcal/mol, slightly higher
than the value calculated by Alabugin for the 1,5-shift
of 1i using a lower level of theory. The barrier for the
1,2-shift is 55.1 kcal/mol, significantly higher than that
for the 1,5-shift, whereas the barrier for ring opening
isomerization is 34.5 kcal/mol, intermediate between the
calculated values for the 1,5- and 1,2-shift.
Assignment of the upfield multiplet to the methylene
protons proximal to the phenyl substituent is consistent
with the expected shielding effect of the twisted 1-phenyl
substituent and is supported by analysis of the NOESY
spectra of 3a. Comparison of spectrum of 3a-d10 with that
of 3a indicates that aromatization of the putative inter-
mediate 3i occurs predominantly via 1,2-deuterium
migration rather than 1,5-deuterium migration. A 85:15
ratio for 1,2- vs 1,5-deuterium migration can be calcu-
lated from the integrated 2.78 and 2.81 ppm peak areas
(Scheme 4b). The same ratio was obtained following
room-temperature irradiation of a glassy solid film sup-
ported on a quartz plate. The NMR spectra of solutions
of 3-d10 irradiated at 195 and 95 K (see Experimental
Section) provide ratios of 75:25 and 55:45 for the mul-
tiplets at 2.78 and 2.81 ppm. The quantum yields for
cyclization of 3 and 3-d10 at 298 K are identical.
Competitive 1,2- vs 1,5-Shifts. The mechanisms
outlined in Scheme 4 are formulated on the assumption
that both 1,2- and 1,5-hydrogen shift products are formed
via the ground state 8a,9-dihydrophenanthrene interme-
diates 2i and 3i. The 1,5-shift can occur via a concerted
symmetry allowed processes, whereas the 1,2-shift prod-
uct requires two sequential 1,2-shifts. The intermediate
formed in the first 1,2-shift is shown in Scheme 4 as a
cyclopropane rather than a 1,3-biradical, based on its
calculated structure (vide infra). Competitive sequential
The structures of 2i and the transition states and
products of 1,5- and 1,2-hydrogen migration are shown
in Scheme 6. Based on the C‚‚‚H‚‚‚C bond distances, the
(14) Rodr´ıguez, D.; Navarro-Vazquez, A.; Castedo, L.; Dominguez,
D.; Saa´, C. J. Org. Chem. 2003, 68, 1938-1946.
(15) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.,
Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.;
Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-Gordon, M.; Replogle,
E. S.; Pople, J. A., Gaussian 98, Revision A.9; Gaussian, Inc.: Pitts-
burgh, PA, 1998.
J. Org. Chem, Vol. 70, No. 25, 2005 10449

Lewis et al.
SCHEME 5. Energies (kcal/mol) Calculated for
Nondeuterated 2a
Alternative Mechanisms. Experimental kinetic data
is not available for highly exothermic cyclohexadiene 1,5-
shift processes, and thus it is possible that DFT calcula-
tions substantially overestimate the barriers for such
processes. A concerted double 1,2-shift in which the
transition state is stabilized by the development of
aromaticity provides a possible alternative to the sequen-
tial 1,2-shift mechanism. We have been unable to locate
a transition state for such a process and are unaware of
any literature precedent. It is possible that both the 1,5-
and 1,2-shift processes occur via tunneling mechanisms
resulting in substantial lowering of the apparent barriers.
The activation parameters, namely, the low activation
energy and preexponential, obtained for the bleaching of
4i (Scheme 3) are consistent with a tunneling process.
Tunneling is traditionally associated with reactions that
occur at low temperature and with large kinetic isotope
effects.17 As an example, Grellmann et al. observed
nonlinear Arrhenius plots and large kinetic isotope effects
(50-100) at low temperatures for sigmatropic hydrogen
shift reactions.19 However, we note that similar 1,2- and
1,5-shift product ratios are obtained for 2-d5 and 3-d10
even though a deuterium shift leads to the major product
for 2-d5 and a hydrogen shift leads to the major product
for 3-d10. Peters and Kim have recently pointed out that
for highly exothermic proton-transfer reactions tunneling
persists at room temperature and can have a negligible
isotope effect, a result in accord with the Lee-Hayes
tunneling theory.20
Another alternative mechanism for the 1,2-shift would
require the presence of a reaction channel for singlet
state decay that leads directly to the 1,2-shift products,
in competition with formation of the thermally equili-
brated intermediates 2i or 3i, which lead predominately
to the 1,5-shift products. This situation might occur either
if the photocyclization processes involves a conical inter-
section for which the 1,2-shift is the major exit channel
at room temperature or if photocyclization occurred via
an adiabatic potential energy surface leading to the
excited-state intermediates (e.g., 2i*) that could undergo
an allowed 1,2-shift in competition with decay to the
ground-state intermediates, 2i.21 Either mechanism would
account for the weak 510 nm transient absorption
spectrum observed for 3.
a (a) Nonadiabatic photocyclization, (b) adiabatic photocycliza-
tion, (c) thermal 1,5-hydrogen shift, (d) thermal 1,2-hydrogen shift,
(e) thermal ring-opening isomerization, and (f) excited state 1,2-
hydrogen shift.
transition state for 1,5-hydrogen migration is early (1.335
and 1.721 Å), whereas the transition state for 1,2-
migration is relatively late (1.473 and 1.228 Å), in accord
with the Hammond-Leffler postulate.16,17 The bond
lengths for the 1,5-shift are in good agreement with those
reported for 1i by Alabugin et al.5 The product of the 1,2-
shift is best described as a cyclopropane rather than a
1,3-biradical, based on the calculated bond lengths (1.50-
1.60 Å for the three cyclopropane bonds).
On the basis of these calculations, the 1,5-shift should
be the predominant reaction pathway for the thermal
isomerization of 2i. Our deuterium labeling results
establish that the 1,5-shift is slightly favored over the
1,2-shift upon irradiation of 3-d10 at 90 K in a MC glass
but that the 1,2-shift predominates for both 2-d5 and 3-d10
at 298 K. The occurrence of a 1,2-deuterium shift and
1,2-hydrogen shift as the major process for 2-d5 and 3-d10,
respectively, is inconsistent with the high barrier calcu-
lated for the 1,2-shift (Scheme 5). In addition, the barriers
calculated for the 1,5-shift of 2i is too high to account for
the rapid bleaching of the UV bands of 2i or 3i in fluid
solution at low temperatures. McCullough and co-work-
ers were successful in detecting the isoindene intermedi-
ate 5i formed upon irradiation of 5 at 203 K (Scheme 7)
both by 1H NMR and chemical trapping, even though the
experimentally determined barrier (Ea ) 14.4 kcal/mol,
A ) 20.5) is smaller than the calculated barrier for 3i.18
The formation of 3i can be observed by means of UV
spectroscopy upon irradiation in a frozen methylcyclo-
hexane glass at 77 K but not in liquid propane solution
at 90 K. From these observations, we are forced to
conclude that these reactions do not proceed by surmount-
ing the calculated barriers shown in Scheme 5.
Adiabatic electrocyclic photochemical reactions have
been observed in cases where the excited state of the
product lies well below that of the reactant.22 Adiabatic
isomerization of 2* to 2i* is estimated to be exothermic
by ca. 16 kcal/mol (Scheme 5, path b) on the basis of the
calculated ground-state energies and singlet energies
estimated from the 0,0 absorption bands. However,
exothermic isomerization is a necessary but not sufficient
condition for the occurrence of adiabatic cyclization.
Further information about these alternative mechanisms
awaits exploration of the excited-state potential energy
surfaces.
(16) The length of the bond being broken is reported first. Leffler,
J. E. Science 1953, 117, 340-341. Hammond, G. S. J. Am. Chem. Soc.
1955, 77, 334-338.
(19) Grellmann, K. H.; Schmitt, U.; Weller, H. Chem. Phys. Lett.
1982, 88, 40-45.
(20) Peters, K. S.; Kim, G. J. Phys. Org. Chem. 2005, 18, 1-8. ()
Lee, S.; Haynes, J. T. Chim. Phys. 1996, 93, 1783-1807.
(21) Klessinger, M.; Michl, J. Excited States and Photochemistry of
Organic Molecules; VCH Publishers: New York, 1995.
(22) Turro, N. J.; McVey, J.; Ramamurthy, V.; Lechtken, P. Angew.
Chem., Int. Ed. Engl. 1979, 18, 572-586.
(17) Isaacs, N. S. Physical Organic Chemistry; 2nd ed.; Longman:
Essex, England, 1995.
(18) Kamal de Fonseka, K.; Manning, C.; McCullough, J. J.;
Yarwood, A. J. J. Am. Chem. Soc. 1977, 99, 8257-8261. () McCullough,
J. J. Acc. Chem. Res. 1980, 13, 270-276.
10450 J. Org. Chem., Vol. 70, No. 25, 2005

Mechanism of 2-Vinylbiphenyl Photocyclization
SCHEME 6
the mixture was stirred for an additional 30 min. Subse-
quently, acetone-d6 (700 mg, 10 mmol) was added slowly and
mixture was stirred for 24 h. After workup, the crude mixture
was chromatographed by column chromatography and elution
with hexane gave 1.6 g (73%) of the Grignard product, R,R-
dimethyl-2-biphenylmethanol-d6, along with traces of biphenyl
and 2-bromobiphenyl. The R,R-dimethyl-2-biphenylmethanol
(500 mg, 2.1 mmol) was dissolved in 50 mL of dry benzene
and refluxed with p-toluenesulfonic acid (25 mg) for 3 h using
a Dean-Stark condenser to remove water. The reaction
mixture was filtered through Celite and further purified by
silica gel chromatography. Elution with hexane afforded 300
mg (70%) of the product as a colorless oil. 1H NMR (CDCl3,
400 MHz): δ 7.31-7.43 (9H, m, aromatic). 13C NMR (CDCl3):
δ 127.0, 127.4, 127.4, 128.3, 129.1, 129.4, 130.4, 139.8, 142.3,
143.0, 146.6. Mass spectrum (relative intensity): M+ 199 (25),
181 (100), 167 (5), 153 (6), 90 (6).
2-Vinyl-1,3-terphenyl-d10 (3-d10). To a stirred solution of
Pd(PPh3)4 (720 mg, 2 mol %), CuI (230 mg, 4 mol %), dry
piperidine (5 mL), and bromobenzene-d5 (5.0 g, 30.9 mmol) in
anhydrous THF (50 mL) was added dropwise trimethylsilyl-
acetylene (5.25 mL, 37 mmol) at room temperature. The
mixture was stirred at 50 °C for 48 h, then quenched with
water, extracted (ether), washed (diluted HCl, brine), dried
(MgSO4), filtered through a short column of silica gel, and
concentrated. The residue was purified by preparative column
chromatography on silica gel (eluent hexane) to obtain 2.8 g
(15.8 mmol, 51%) of trimethyl(phenylethynyl)silane-d5. This
was dissolved in methanol (10 mL), and anhydrous K2CO3 was
added in one portion. The reaction mixture was stirred for 30
min at room temperature, then water was added, and the
mixture was extracted with pentane, washed (water, brine),
dried (MgSO4), filtered, and concentrated under ambient
pressure. The resulting crude product was added dropwise to
a stirred mixture of PdCl2(PPh3)2 (100 mg), CuI (200 mg), dry
diethylamine (5 mL), and 1 M solution of vinylbromide in THF
(20 mL, 20 mmol). The mixture was stirred at room temper-
ature for 2 h, then poured into ice-cold 5% HCl (150 mL),
extracted with ether, washed (water, brine), dried (MgSO4),
filtered, and evaporated in a vacuum from an ice-bath. The
residue was purified by preparative column chromatography
on silica gel (eluent pentane) to afford 1.5 g (11 mmol, 72%) of
but-3-en-1-ylbenzene-d5. Benzannulation was carried out in
two parallel runs. Two 5 mL minireactors equipped with mini-
inert valves were loaded with but-3-en-1-ylbenzene-d5 (750 mg,
5.6 mmol), Pd(PPh3)4 (200 mg, 0.17 mmol, 3 mol %), and
SCHEME 7
Concluding Remarks
The results of deuterium labeling studies of the pho-
tochemical rearrangements of the 2-vinylbiphenyl de-
rivatives 2-d5 and 3-d10 establish that product formation
occurs predominantly via 1,2-deuterium and 1,2-hydro-
gen shifts, respectively. Product formation via 1,5-
hydrogen and deuterium shifts account for the minor
reaction products at room temperature but increase in
importance with decreasing temperature in the case of
3-d10. Theoretical investigation of the thermal activation
barriers for 1,2- and 1,5-shifts are consistent with a
significantly lower barrier for the 1,5-shift, as expected
for the thermally allowed sigmatropic rearrangement.
The preferred formation of the products of 1,2- vs 1,5-
hydrogen or deuterium migration appears to be incom-
patible with competitive ground-state isomerization path-
ways. Alternative pathways involving direct formation
of the 1,2-migration products from excited intermediates
need to be considered.
Experimental Section
Syntheses and characterization of the nondeuterated vinyl-
biphenyls 2 and 3 and their photochemical conversion to the
dihydrophenanthrenes 2a and 3a have been reported.7,23
2-Isopropenylbiphenyl-d5 (2-d5). The deuterium-labeled
2-isopropenylbiphenyl was prepared using a method similar
to that for the corresponding protiated derivative 2.7 A solution
of MgCl2 (1.2 g) and potassium metal (1.5 g) was refluxed in
75 mL of dry THF under dry nitrogen for 3 h. To this ash-
colored mixture was added 2-bromobiphenyl (2.33 g, 10 mmol)
slowly in 20 mL of dry THF. After the addition was complete,
(23) Lewis, F. D.; Crompton, E. M.; Sajimon, M. C.; Gevorgyan, V.;
Rubin, M. Photochem. Photobiol. 2005, ASAP.
J. Org. Chem, Vol. 70, No. 25, 2005 10451

Lewis et al.
anhydrous THF (3 mL). The reaction mixture was stirred at
100 °C for 48 h and then cooled, and the contents of the two
reactors were combined. The mixture was filtered trough a
short column of silica gel (CH2Cl2 eluent) and concentrated.
The residue was purified by preparative column chromatog-
raphy on silica gel (hexane/CH2Cl2 10:1 eluent) to afford
2-vinyl-1,3-terphenyl-d10. Yield 780 mg (2.9 mmol, 52%). Mp
107-108 °C. 1H NMR (CDCl3, 400 MHz): δ 4.60 (1H, d), 5.02
(1H, d), 6.44 (1H, m), 7.26 (2H, m, aromatic), 7.35 (2H, m,
aromatic). 13C NMR (CDCl3): δ 121.7, 126.9, 127.4, 127.6,
127.9, 129.9, 134.9, 135.1, 141.8, 142.3. Mass spectrum (rela-
tive intensity): M+ 266 (100), 259 (5), 248 (7), 184 (10), 170
(5), 123 (5).
Photolysis of 2-d5. Room-temperature irradiation was
performed in a degassed solution of cyclohexane (30 mg in 100
mL) using a quartz vessel with 254 nm excitation provided by
a chamber reactor. The reaction was monitored by GC-MS
to the point of complete conversion. The irradiated solution
was filtered through silica gel, and the solvent was removed
under reduced pressure to afford the photoproduct in quan-
titative yield. 2a-d5: 1H NMR (CDCl3, 400 MHz): δ 2.75
(corresponding to the methylene proton derived from a 1,5-
shift), 3.09 (corresponding to the methylene proton derived
from a double 1,2-shift), 7.21-7.76 (8H, m, aromatic). 13C NMR
(CDCl3): δ 27.1, 32.7, 123.8, 124.1, 126.9, 127.0, 127.1, 127.4,
127.7, 127.9, 128.3, 129.0, 129.1, 129.3. Mass spectrum (rela-
tive intensity): M+ 199 (40), 181 (100), 153 (5), 90 (8). Analysis
of the integrated NMR spectrum shows that the mixture
contains 90% of 2a-d5 derived from a 1,2-shift and 10% from
a 1,5-shift.
Photolysis of 3-d10. Room-temperature solution irradiation
was performed as described for 2-d5. Irradiation at was also
performed on glassy solid film of 3-d10 prepared by evaporation
of a dichloromethane solution on a quartz plate. Irradiation
at 195 K was performed in cyclohexane solution containing
added dry ice. Irradiation at ca. 85 K was conducted in liquid
propane using a nitrogen-cooled cryostat. To increase solubility
of the reactant in propane, a thin layer of the reactant was
deposited in a quartz cuvette by evaporation of a dichlo-
romethane solution. To the dried cuvette equipped with a small
stir bar, liquid propane was condensed at low temperature and
the resulting mixture stirred for about 30 min. This solution
was irradiated with constant stirring. The propane was
allowed to evaporate and the residue dissolved in cyclohexane.
Reactions were monitored to the point of complete conversion
by GC-MS analysis. The irradiated solutions were filtered
through silica gel, and the solvent was removed under reduced
pressure to afford the photoproduct. Room-temperature ir-
radiation of 30 mg of 3-d10 in 100 mL cyclohexane afforded
3a-d10 in quantitative yield. 1H NMR (CDCl3, 400 MHz): δ 2.75
(2H, d), 2.81 (1H, d), 7.26 (1H, d, aromatic), 7.37 (1H, t,
aromatic), 7.79 (1H, d, aromatic). 13C NMR (CDCl3): δ 26.6,
29.1, 121.7, 123.6, 126.7, 127.0, 129.4, 129.9, 135.1, 137.6,
141.3. Mass spectrum (relative intensity): M+ 266 (60), 259
(10), 248 (25), 184 (100), 155 (8). Ratios of isomeric photoprod-
1
ucts were determined by analysis of the integrated H NMR
peak areas for the corresponding aliphatic protons. The
product obtained from irradiation at room-temperature either
in solution or as a glassy solid contains 85% of 3-d10 derived
from a 1,2-shift and 15% via a 1,5-shift. The product obtained
from irradiation at 195 K contains 75% of 3-d10 derived from
a 1,2-shift and 25% via a 1,5-shift. The product obtained from
irradiation at 85 K contains 45% of 3-d10 derived from a 1,2-
shift and 55% via a 1,5-shift.
Transient Absorption Spectroscopy. The transient ab-
sorption measurements were performed at the Ohio Labora-
tory for Kinetic Spectroscopy at Bowling Green State Univer-
sity, and the experimental setup has been detailed elsewhere.24
In brief, the output of a Ti:sapphire laser (fwhm ) 150 fs) was
steered into a third harmonic generator to obtain the 267 nm
excitation wavelength. Sample solutions were prepared to have
an absorbance of 0.7-1.0 at the excitation wavelength and
were used without deaeration. The sample flow-through cell
had an optical path of 2 mm and was connected to a solution
reservoir and flow system. All measurements were conducted
at room temperature, 22 ( 2 °C.
Computational Methods. Ab initio calculations were
performed by using the GAUSSIAN 98 program.15 Geometry
optimizations were carried out using B3LYP25 density function
with 6-31G** basis set, and energies were calculated at
B3LYP/6-311G** level. Transition states were located by using
the synchronous transit-guided quasi-newton method26 at
B3LYP/6-31G* level. All stationary points were identified as
minima or transition states by vibrational analysis. The
transition states were characterized by the presence of one
negative vibration frequency and confirmed by intrinsic reac-
tion coordination analyses.
Acknowledgment. Funding for this project was
provided by NSF grants CHE-0400596 (F.D.L.) and
CHE-0354613 (V.G.).
Supporting Information Available: Structures and Car-
tesian coordinates for reactant, intermediates, and transition
states. 1H NMR and 13C NMR spectra of reactants and
products. This material is available free of charge via the
Internet at http://pubs.acs.org.
JO051730Y
(24) Shah, B. K.; Rodgers, M. A. J.; Neckers, D. C. J. Phys. Chem.
A 2004, 108, 5926-5931.
(25) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785-789.
Becke, A. D. Phys. Rev. A 1988, 38, 3098-3100. Becke, A. D. J. Chem.
Phys. 1993, 98, 5648-5652.
(26) Peng, C.; Schlegel, H. B. Isr. J. Chem. 1993, 33, 449-454.
10452 J. Org. Chem., Vol. 70, No. 25, 2005